Research on Interpolation Algorithms of Digital Storage Oscilloscope

Abstract

Digital storage oscilloscope can only collect a few sample points of signal because of insufficient real-time sampling rate when it works in fast time base gear. In order to reconstruct the signal, interpolation must be carried out. This paper introduces the methods of piecewise linear interpolation and cubic spline interpolation in detail, and compares the recovery results of several different waveforms by programming and simulation, including cubic spline interpolation, piecewise linear interpolation, cubic interpolation and nearest neighbor interpolation, and calculates and analyses the related mean square error. Considering the unknowability and complexity of the actual signal to be measured, and weighing the accuracy and speed, when the number of points to be interpolated is less than 100, the cubic spline interpolation method is used; when the number of points to be interpolated is not less than 100, the cubic interpolation method is used. The simulation results of MATLAB show that all the performance indexes of the design results meet the specified requirements, and the design process is simple and feasible.